The rush of activity among London's auction houses in the first few weeks of summer has long been a familiar occurrence that persists even today. However, this intense seasonal concentration of sales was not always so. This paper draws on quantitative methods to explore the gradual emergence of a tightly scheduled auction season in London at the turn of the nineteenth century, focusing on the sale of paintings. By analysing historical art auction catalogue data, the paper traces the ways in which this shift varied across different segments of the auction market, as well as between individual auction houses. As our study shows, the temporal clustering of painting auctions had specific business advantages, but it also played a key role in enhancing the social import of these auctions, demarcating an annual, weeks-long “event” looked to with anticipation and excitement by auctioneers and buyers alike.

Recent literature has highlighted the importance of spatial structures to the evolution of the nineteenth-century London art market. The geography of commercial galleries and auction houses was tightly interwoven with that of exhibition society halls. The Royal Academy was a prime centre of artistic gravity, with galleries and auction halls opening near its first seat on Pall Mall in 1768, and many following its path to Somerset House in 1780 and to Burlington House in 1867.¹ Exhibiting institutions and commercial galleries established a mutually beneficial relationship cemented by spatial proximity to each other, as well as to the fashionable retail shops of the West End.²

Yet there was also a dynamic temporal dimension to the social world of nineteenth-century London. It has been argued that a prime period for art auctions in London in the nineteenth century fell between late May and early June, coinciding with the yearly influx of “Society” from their country houses.³ This was not always the case, however. While previously assumed to be a static influence, close study reveals that the seasonal structure of the art auction market had a history all its own.

Responding to Pamela Fletcher’s and Anne Helmreich’s call for the use of large data sets over case studies alone in characterizing the development of the London art market, this study draws on a database of painting auction records between 1780 and 1835 in order to tease out the temporal structures in the early flourishing of art auctions in London, and test their relationship to the schedule of the Royal Academy and the larger “Season” of London society.⁴ We demonstrate that highly concentrated auction activity in the early summer only developed in the early nineteenth century, and did not apply to all types of artworks, nor was it followed uniformly by all auction houses in London. Moreover, this move towards a more intense auction season should not be interpreted merely as a result of auctioneers’, exhibitors’, and buyers’ financial interests. This temporal concentration may have acted as a focusing lens that compounded the social import of the London art market in this period, setting apart fine art auctions not just in place, but also in time.

One of the most comprehensive stores of information about the London art market in the late eighteenth and early nineteenth centuries can be found in the Getty Provenance Index Sales Catalog Database, a project of the Getty Research Institute (GRI). The Sales Catalog Database contains structured descriptions of sales catalogues listing works of art for public auction in major European countries from the seventeenth through the nineteenth centuries. A subset of this endeavour, the British Sales Catalogs project, was developed from an extensive search of hundreds of libraries and archives in order to locate surviving catalogues of British sales, identifying almost 9,000 surviving catalogues dating between 1681 and 1850. The Getty has been gradually indexing these catalogues, entering details of the sales of individual artworks into a searchable database.⁵ While the project is still ongoing, the most completely described catalogues of this group span the years between 1780 and 1835, with almost 95 percent of known catalogues from this period having been indexed at the level of the individual artwork (fig. 1).⁶

Like any archive, this database is not a perfect representation of all historical auctions as they actually occurred; we face that intractable historiographic problem in tandem with all historians. There are inevitably auctions that occurred during this period whose catalogues, by random accident, have not survived. That said, there are also sources of systematic bias: particular classes of sales whose records we can, with specific reasons, claim are disproportionately absent from the corpus of surviving auction catalogues. For example, analyses based on the Getty data cannot speak to the patterns of private sales that went unrecorded in published catalogues. Likewise, catalogues with a very small run, or from particularly minor houses, that have not survived into today’s institutional collections may also be under-represented. The database also has only limited coverage of auctions of sculpture, works on paper, and decorative arts. With these biases in mind, readers should be aware that the claims in this paper will be restricted to the trends in scheduling of public auctions of paintings.

As of August 2015, the British Sales Catalogs database contains 361,112 entries for individual painting lots in auctions taking place in London. Of all these records, 316,633 (88 percent) have dates falling between 1780 and 1835.⁷ Of those records, 230,365 (64 percent of the original total) are listed with a transaction price.⁸ It is this last group of records that comprises the working data for our analysis.

Only about 60 percent of these priced objects were sold, with around 25 percent listed as “bought in” (fig. 2). The prices that were recorded for these unsold objects may be the unmet reserve price, or simply the price at which bidding stopped. Because we are interested in auctioneers’ strategies in scheduling major sales, we include both works that have been marked as “sold” as well as those “bought in” or “withdrawn” where the listed price is either a reserve or an estimate.⁹

Regarding prices: although we are not concerned with measuring art price fluctuations over the years as such, it is nonetheless important to distinguish between auction patterns of the mostly costly paintings versus those sales of less expensive ones, as different segments of the auction market may well have followed different seasonal patterns.¹⁰ Historical inflation and deflation make direct comparison of price values from one decade to the next difficult.¹¹ For example, a £6 painting in 1780 would belong to the top 20 percent of all works from the same year, while the same price in 1835 would barely break the median price of that year. Therefore, this analysis classifies records across the entire period of study into one of five bins, or quintiles, compared to other records within the same year of sale. If a work was particularly cheap in its own year of sale, then we group it alongside artworks that were also inexpensive in their own respective years of sale; likewise for works that were particularly expensive. For example, a sale that fell into the lowest 20 percent of prices (the first quintile) in 1780, and a sale that fell in the first quintile in 1830, can be considered together, as could the highest 20 percent of works in the fifth quintile. Sorting works into quintiles relative to their own year of sale allows us to generally divide “cheap” artworks from the most “expensive” artworks across the entire period of study, and thus ask how their respective seasonal patterns may have differed.

One measure of seasonality in the market can be found by considering how tightly or loosely the busiest auction days cluster in a given year. Are the days with the most auctions concentrated within the span of a few weeks? Or are they scattered throughout the entire year? For each year between 1780 and 1835, we identified the top seven days as measured by the number of lots being sold.¹² To characterize how this annual “spread” changed over time, we measured the coefficient of variation (CV) of the top seven days’ locations within the calendar year.¹³ A high CV would indicate that these peak days were relatively scattered throughout the year, while a low CV would indicate that these days were more tightly clustered together (see fig. 6 for a visualization of this procedure). It then becomes possible to track year by year, quintile by quintile, and auction house by auction house, whether this CV remained the same, showing no indication of a concentrating season, or if it instead decreased, indicating that the top auction days in a given year were occurring closer and closer together.

Figure 7 illustrates the result of these calculations, split into rows based on the auction house (all sales, sales just by Christie’s, and sales by others), and split into columns based on the price quintile of the artwork being sold. Each plot shows a line of best fit that helps to illustrate whether or not the yearly CV remained the same, increased, or decreased overall between 1780 and 1835. We find that the timing of the “peak auction season” for the cheapest artworks (those in the first to fourth quintiles) did not coalesce significantly over the course of this study period. However, the seasonal spread of the top sales days did tighten significantly for the most expensive paintings in the fifth quintile. Even more intriguing, Christie’s appears to be the strongest driver of this trend towards a tighter schedule. Because Christie’s accounts for 36 percent of the sales records, its auctioneering strategies have a disproportionate effect on measurements of all auctions combined. When disaggregated, Christie’s sales show a much greater increase in seasonal concentration than do sales by other auction houses represented in the data. For example, the CV of Christie’s top sales days decreased, on average, by 71 percent between 1780 and 1835, while the CV of houses other than Christie’s only decreased by 10 percent.

Around which dates did different auction houses organize their busiest sale days? It is useful to focus on the scheduling strategies of the three most active auction houses in this data set: Christie’s, Edward Foster, and Harry Phillips. Figure 8 plots the top auction days for fifth quintile (most expensive) artworks from these houses, annotated with the start and end dates of the Royal Academy exhibition in each year.¹⁴ In 1780, Christie’s scheduled its largest sales days as early as January and as late as July. While Christie’s would continue to schedule a handful of large sales in the earliest months of the year, after 1800 the house began to concentrate the bulk of sales within the season defined by the dates of the Royal Academy exhibition. (It is also notable that the Royal Academy progressively lengthened its exhibition in this period, from just one month in 1780 to over three months in 1835.) The second largest auctioneer of paintings in this data set, Harry Phillips, similarly distributed large sales between January and July during their earliest years of sales shortly before 1800. Like Christie’s, Phillips increasingly scheduled the highest-volume sales days during or directly before the Royal Academy exhibition. A curious exception, however, was Edward Foster, a later entrant to the market, whose first public sales took place in 1812. While, like Christie’s and Phillips, Foster did hold larger sales during the Royal Academy exhibition, he also frequently held top sales days during the late summer and fall months—a time of the year during which most other auction houses went dormant. Counting up the top sales days for each of these three houses, 23 percent of Foster’s top days happen after 1 August, while only 4 percent of Christie’s and Phillips’s do so. Foster’s expansion into the fall months presents an intriguing parallel with another abnormal pattern of this auction house: it scheduled most of its paintings auctions in the middle of the week (fig. 9), while Christie’s and Phillips overwhelmingly favoured Friday and Saturday sales. 

This analysis sheds light upon three under-studied points about the London auction season: that a concentrated sales period emerged only gradually around the turn of the nineteenth century, that it applied far more to the upper echelon of painting auctions than to sales of middle- or low-end works, and that not all auctioneers devoted their sole attention to this season.

As much as the London art world overlapped with the geography of luxury retailers in the West End, so too was it attuned to the temporal structure of the London Season.¹⁵ Commercial galleries and artists were keenly aware of the timing of Royal Academy exhibitions, and the seasonal concentration compounded with the geographic concentration of the market to engender both competition as well as cooperative interaction. Figure 8 shows how the Royal Academy acted as a temporal attractor for high-end art auctions, as Christie’s and other auction houses increasingly scheduled their biggest sales days to take place during the Academy’s summer exhibition. Established as a focal point, and in some regards the kickoff, of the London Season, the Royal Academy exhibitions and artists were seen by many institutions as a means to draw audiences already inclined towards the viewing of art. Alderman John Boydell’s Shakespeare Gallery seized on this predisposition as a means to promote English artistic production and patronage as a patriotic act, as well as to drum up interest in the purchase of engravings after works in the gallery.¹⁶ The Shakespeare Gallery opened to great fanfare on 4 May 1789 in a Neoclassical structure at 52 Pall Mall, blocks away from the rooms at 125 Pall Mall which had been occupied by Christie’s since 1770. Other thematic galleries later sought to capture a share of the public’s desire for art: Thomas Macklin’s Gallery of the Poets, which was open between 1788 and 1797, and Henry Fuseli’s short-lived Gallery of the Miltonic Sublime, which welcomed visitors in 1799 and 1800.¹⁷ In a true example of the London art world cannibalizing itself, when the Shakespeare Gallery failed and its entire holdings were sold off by lottery in January 1805, 52 Pall Mall itself was leased to the newly created British Institution. Attempting to capitalize on public interest and excitement leading up to the Academy’s annual exhibitions, the British Institution exhibition began in mid-January and its closing dates overlapped with the opening days of the Academy exhibition. While some Academicians were wary of decreased exhibition attendance and proceeds, others praised the British Institution in its efforts to encourage interest in contemporary English artists.¹⁸ Certainly the twenty-eight full Academicians who sent works to the British Institution welcomed the friendly competition.¹⁹ Other artists, most prominently John Singleton Copley, also took advantage of the popularity of the Royal Academy exhibition as a temporal cue for the display of individual or small groupings of history paintings.²⁰

By scheduling a battery of major sales within the few weeks surrounding the social season’s major focal point, auction houses were able to maximize publicity for their sales and concentrate the attention of potential buyers, many of whom belonged to the landed gentry who spent the majority of the year in country estates away from London. However, for all its recognized influence on the schedules of other exhibiting organizations and individuals, we find the Royal Academy was not a consistent temporal attractor for painting auctions in the late eighteenth century. As figure 8 illustrates, not until several decades after the founding of the Royal Academy did the major London auctioneers begin to schedule their highest-volume sales days during the annual exhibition.²¹ William Roberts’s 1897 history of Christie’s firm corroborates the evidence that James Christie, in particular, maintained an unpredictable auction schedule in the late eighteenth century:

Mr. Christie’s picture sales sometimes ran in fits and starts. Several would occur within a few weeks of one another, and then cease for months. After the Colebrooke dispersal, for example, no more noteworthy picture sales occurred until December 14th. From the catalogues it would appear that there was no such institution as a vacation invented, for auctions were held throughout August, September and October—whenever, in fact, they were wanted or could be made up.²²

Our analysis supports Roberts’s characterization of Christie’s scattered auction schedule in the 1780s and 1790s. Moreover, our results also suggest that the house began to follow a more regular and concentrated summer auction schedule after 1800. It may not be a coincidence that this change came after James Christie died in 1803 and the house passed to his eldest son, also named James. The younger Christie seems to have been a far more responsible and capable businessman than his father, who had a reputation for difficult business dealings and was notorious for overcharging clients for fees and for extremely late payment from auction proceeds.²³ While other London auction houses also gradually shifted their top auction days more closely together, it was Christie’s that made the most striking shift.

The seasonal influence of the Royal Academy exhibition and other Society events on auction schedules was not necessarily a one-way relationship, however. Early exhibitions of English art, such as those at the Foundling Hospital beginning in the 1740s, and a number organized by the Society of Artists of Great Britain from 1760 onwards, were in turn responses to a growing interest in art stoked by access to art auction previews.²⁴ It is possible that the Royal Academy also responded to the growing volume of art auctions over this period by gradually lengthening its annual exhibition while shifting its opening day to late April, and then to early May. In turn, Christie’s and Harry Phillips themselves also appear to have extended their own high-volume seasons to follow the lengthening Royal Academy exhibition. Both the Royal Academy and the auction houses may also have been gradually adapting to third-party factors as well. In the early nineteenth century, the majority of Parliament sessions ended after the start of July. This was consistently later in the year than session closing dates in the mid-eighteenth century, when as many ended in April or March as did in June or July.²⁵ The art market, along with the rest of high London society, may still have been gradually adjusting its schedule to Parliament’s new pattern.

Another factor that may have incentivized auction houses to concentrate high-end sales within a shorter period of time was the increasing volume of paintings being auctioned. The overall growth in the number of recorded auctions per year (fig. 5) over the course of this period may well have been sparked by political events, as continental aristocracy displaced by revolution and war in the eighteenth and nineteenth centuries were often forced to part with their art collections. Frequently these émigrés found their way to England, as did their art. From 1789 onwards, sales of French collections ballooned, as did shipments from Spain and the Netherlands between 1798 and 1810.²⁶ This increasing volume may have been an additional incentive for auction houses to concentrate their largest sales within a shorter, and thus more efficient, schedule.

Seasonal concentration may have had more than a simple economic motivation, however. The centrality of these art institutions to high-end social life in London is well understood. Auctioneers like Christie’s and Sotheby’s carefully constructed the rhetoric of their auction advertisements and catalogues in order to promote their sales as venues where both fine art, but also social cachet, could be obtained and displayed.²⁷ The concentrated geography of the London art market not only optimized physical foot traffic, but also lent a fashionable West End imprimatur to the art market, an association that mutually benefitted both public exhibitors and private galleries.²⁸ Private viewing days at auction houses could be highly exclusive indeed. Roberts writes:

A great feature of the sales at Christie’s at the latter part of the last century was the private view day. This was a fashionable lounge where persons of distinction congregated in great numbers. During the season, when any remarkable collections were on view, occasional evening receptions took place: the great room was then lighted up, and persons of quality attended in such large numbers that an official from the Opera was stationed at the entrance to prevent the intrusion of those not belonging to the fashionable world.²⁹

James Gillray caricatured the high society audiences of auctions in A Peep at Christie’s;—or—Tally-ho, & his Nimeny-pimmeny, taking the Morning Lounge (fig. 10), showing Edward Smith-Stanley, twelfth Earl of Derby, and the actress Elizabeth Farren viewing works during an auction preview day, while three other attendees in outfits of varying levels of absurdity dominate the background.³⁰ The caricaturist also lampooned the bombastic rhetoric of Christie’s advertisements and catalogues (and the credulous buyers who succumbed to them) in Mæcenas, in pursuit of the fine-arts (fig. 11), in which Gillray satirically pictures George Granville Leveson-Garner, Marquess of Stafford, a notable patron of the arts, being drawn into a Christie’s sale outlandishly advertised with a “Catalogue of 800 Capital Pictures to be Sold by Mr. Christie in Pall Mall, February 1st, 1808”. That there was, in fact, no prominent sale on that date in 1808 points to this Maecenas’ gullibility.³¹ Based on our analysis of auction dates, we might also note that he was less likely to find auctions of truly capital pictures in London while there was still snow on the ground.

Landscape painter Joseph Farington’s daily diary, kept between 1793 and 1821, provides exceptional insight into the minutiae of life for members of English high society. A founding member of the Royal Academy and distant relative, by marriage, of politician and art historian Horace Walpole, fourth Earl of Orford, Farington was one of the more successful English artists of his day in terms of leveraging his artistic insight toward increased social status. In 1806 he recorded an experience of some of the art-viewing opportunities available around Pall Mall during the Season. On Friday, 25 April he spent the day in the company of famous collector Sir George Beaumont, seventh Baronet, during which the pair visited Christie’s to view Dutch and Flemish paintings, followed by a visit to another gallery to look at works by Poussin and Rubens, and ending at the inaugural British Institution exhibition, where they joined a discussion on the new institution’s challenge to the Royal Academy, whose exhibition would open exactly a week later.³²

These sales showcased elite consumption in a very public way, to a wide range of ends, as Gillray’s print of the Earl of Derby and Elizabeth Farren so vividly depicts. Many auction attendees were there to participate in the spectacle by demonstrating their refined taste through the purchase of art and the construction of new collections. Others viewed collecting as a patriotic venture, demonstrating conspicuous cultural patriotism through the purchase of works by contemporary English artists.³³ Like the Royal Academy exhibition itself, auctions of prestigious collections were events at which one could see and be seen by “Society”.³⁴ These events became venues for the performance and construction of taste and refinement. In 1808, William Henry Pyne quipped that “those who might think it necessary to appear to have, what nature had denied them, taste and judgment”, relied on the dealers and auction rooms of London to address their lack of discrimination, spotlighting the performative aspect of auction attendance.³⁵

An emergent seasonal structure may have helped to further demarcate exceptional auction periods from day-to-day commerce, elevating art auctions from a series of interchangeable incidents to an annual, weeks-long “event” looked to with anticipation and excitement by auctioneers and buyers alike. Arjun Appadurai’s conception of auctions as “tournaments of value” is particularly apt here. Appadurai characterizes these tournaments as periodic events that are as much a way to signal social status as to exchange material goods, and, as such, are socially distinguished from everyday commercial activity.³⁶ Appadurai suggests that it is the sport-like rituals of the auction room that form this distinction, and our analysis demonstrates how the dimension of time is relevant to this interpretation. By fitting their most prestigious auction days into an ever-tighter schedule, houses like Christie’s could concentrate wealth and social cachet within a tightly defined place and time. This was all the more true between 1780 and 1840, when auction rooms were largely filled by retail buyers, and had not yet been dominated by professional dealers buying and selling their own inventories.³⁷ Shortening and intensifying the season for the highest-end auctions not only served a business function, but also enhanced their social import.

For all the incentives for auction houses to tightly cluster almost all their high-end sales during the height of the London social season, it is also important to recognize a notable exception. We found that the third largest auction house represented in these data by sales volume, Edward Foster’s, remarkably scheduled a significant number of its busiest sale days of high-value artworks well into late summer, and even the fall months; by the 1810s, as we have shown, this stands out as an abnormal practice for London auctioneers. This behaviour may be understood as a localized instance of the niche-seeking practice that national markets engaged in on the continental scale. In his broader analysis of international auctions between 1801 and 1820, Christian Huemer has shown that major European art centres appear to have established their own regional auction seasons; England favouring the late spring, France peaking in November, and smaller markets in Belgium and the Netherlands focusing sales in the late summer, when the major markets in London and Paris were largely quiet. Huemer posits that this time-shifting was born of necessity, allowing agents to attend all the year’s important sales without having to be in two places at once.³⁸ It is possible that Edward Foster found similar success in offering some significant sales in the London off-season as a supplement to its regular offerings at the height of the early summer auction season. These later sales could open a venue to those buyers and sellers who would otherwise have had to wait until the next year to offer their works at the biggest sales at Christie’s.

Likewise, it should be noted that Foster was also the only firm we found that appears to have conducted a large portion of its sales midweek (particularly on Wednesday and Thursday—see fig. 9), perhaps taking advantage of the weekly vacuum left by other houses, Christie’s in particular, that favoured Friday and Saturday. While scheduling sales in a short seasonal timeline may have been beneficial for many firms looking to bring together buyers and sellers in a socially charged environment, Foster clearly strove to carve out a unique position even in the midst of the unavoidable peak season, scheduling around behemoths like Christie’s while still taking advantage of the rich field of potential buyers gathered in one season and place.

These results raise some new questions that deserve continued research. To what extent are these same changing scheduling strategies reflected in newspaper advertisements by auction houses? If advertisements had a better survival rate than auction catalogues, particularly for smaller or lower-end sales, then they may offer a fruitful source for checking the representativeness of the Getty’s auction data at different periods in time. Future work might also investigate whether scheduling patterns were affected by particular genres of artwork, or the nationalities of the artists who painted them. Much like the spatial structure defined by physical viewing spaces, the temporal structures defined by auction houses, exhibiting institutions, and the larger social Season played an active role in reshaping the London art market in this period.

For their assistance in providing crucial data and context, the authors would like to thank Christian Huemer, Ruth Cuadra, and Eric Hormell of the Getty Research Institute, and Royal Academy of Arts archivist Mark Pomeroy. William L. Pressly and Martina Droth provided valuable comments, and the work of Tom Scutt on translating our data visualizations for online presentation is most appreciated.